Role of AI chatbots in education: systematic literature review

Labadze, L., Grigolia, M., & Machaidze, L. (2023). Role of AI chatbots in education: systematic literature review. International Journal of Educational Technology in Higher Education, 20(1), 56.

Overview

The article systematically reviews the role of AI chatbots in education, analyzing their benefits, challenges, and potential limitations. It examines the integration of chatbots from the perspectives of students and educators, highlighting key advantages and concerns based on 67 selected studies.

Research Findings Summary

1. Student Perspective:

• Homework and Study Assistance: Chatbots provide detailed feedback on assignments and assist with homework questions, enhancing students’ understanding and preparation for exams.
• Personalized Learning: AI chatbots offer individualized guidance and adapt teaching strategies to each student’s needs, making learning more engaging and personalized.
• Skills Development: Chatbots help develop writing, problem-solving, and discussion skills by providing suggestions, step-by-step solutions, and real-time feedback.

2. Educator Perspective:

• Time-Saving Assistance: AI chatbots assist with routine tasks like scheduling and grading, allowing educators to focus more on instructional planning and student engagement.
• Improved Pedagogy: Educators use chatbots to enhance instructional methods and provide personalized support, generating tailored educational content and activities.

3. Concerns:

• Reliability and Accuracy: There are concerns about the accuracy of information provided by chatbots and the potential for students to rely excessively on AI-generated content, impacting critical thinking and problem-solving skills.
• Ethical Considerations: Issues related to data privacy, security, and ethical use of AI in educational settings are significant concerns that need to be addressed.

A few thoughts

While the study found some results, you can already see the weird positioning the ferver of AI titled and tagged literature is having on these kinds of studies. There’s also an issue of restricting studies to recent years. I recall through most of my undergrad years that the rule was, unless it was a really impactful work (like Dewey or Bloom) that you should focus on research that’s no older than five years. No wonder we always forget what we already know, especially if that’s widely applied to systematic reviews. ChatBots in education have a much longer history and would have provided much better insights overtime than much of what’s presented here.

One example is AutoTutor, which originated in the 1990s and would have fit a definition of AI at the time (insert lament here of the dominant definition of AI being shaped by giant companies in the past few years). AutoTutor had legitimately interesting designs and worked on technology that is quite different from the predictive text-style chatbots we have presented in this paper. I’ll keep my eye out for other reviews that are more inclusive and add this one to the examples from recent years pile.